A target is a physical vapor deposition (PVD) source that is deposited in a PVD tool. The target is depleted by plasma bombardment to sputter coat semiconductor wafers. A target needs to be replaced near the end of its actual target lifetime.
FIG. 1 discloses a graph (100) of an actual target lifetime in KWH of a target being used and depleted in a PVD tool versus the accumulating hours of deposit time, (T). The actual target lifetime is expressed in kilowatt hours (KWH), and depends upon the power rate of the PVD tool and the hours of deposit time for the target in the tool. When the actual target lifetime of graph (100) attains the estimated target lifetime of value (B) in KWH, the target has been depleted to the extent that the target must be replaced after accumulating the hours of deposit time, value (D).
The PVD tool is controlled by a computer to operate a database that records the hours of deposit time (T) for the target in the tool. The KWH power rate of the tool is known, and is recorded in the database. FIG. 1 discloses a reported graph (102) generated by the database. The reported graph (102) reports estimated changes in the estimated target lifetime (KWH) of the target for the accumulating hours of deposit time (T) in the tool. However, the actual hours of accumulating deposit time for the target in the tool can exceed the hours recorded in the database. Such an error occurs when some of the hours of deposit time are unrecorded. Further, such an error occurs due to unexpected equipment system conditions that make the hours of deposit time (T), as reported by the tool, to appear less than the actual hours of deposit time. Thus, the reported graph (102) in FIG. 1 discloses that the reported estimated lifetime change, as reported by the tool, is less than the actual lifetime change of the target (104).
The reported target lifetime remains below the value (B) during the time period from the time (D) to a time (E). The time (E) is past the time (D) corresponding to the actual target lifetime, value (B). Thus, the reported target lifetime, as reported by the tool, will not indicate a trend toward a target overrun that occurs from the time (D) to the time (E). During a target overrun, wafer scraps are produced by the tool. The tool cannot indicate whether a target overrun would be the source of such wafer scraps. Thus, time must be spent to search for the cause of such wafer scraps.
The data reported by the tool would not accurately indicate a trend for when the target in the tool exceeds its actual target lifetime (B). Continuing to operate the tool with a target that has exceeded its target lifetime (B) is a condition referred to as a target overrun, resulting in fabrication of wafer scraps.
For the above reasons, there is a need for the tool to record an estimated target lifetime in a manner that will detect a trend toward target overrun.